Static Headspace-Gas Chromatography-Ion Mobility Spectrometry (SHS-GC-IMS): Its Novel Application to Delineate Quality Grading, Oxidative Alteration, and Ageing Effects in New Zealand Wines

Abstract

e aroma of wine constitutes a crucial aspect of its organoleptic quality, and hence directly links to consumer preference. Also, wine aroma can be easily altered should the wine be subject to oxidative environments or prolonged storage periods. Such characteristics of wine aroma therefore necessitate that it be monitored and evaluated constantly during the winemaking, and short- or long-term storage stages. Despite the importance of wine aroma, commercial wineries usually rely on winemakers to give subjective judgements and verdicts, whereas the appropriate equipment and protocols to perform instrumental analyses are o entimes unavailable. Although sophisticated instruments such as gas chromatography-mass spectrometry (GCMS) are widely used to analyse wine aroma, they are, in most cases, unsuitable for commercial wineries due to large investment and operational costs. In the current PhD work, an analytical instrument named static headspace-gas chromatography- ion mobility spectrometry (SHS-GC-IMS) is applied in studying wine aroma. Firstly, a method was speci cally developed and validated using SHS-GC-IMS for both qualitative and quantitative analyses of wine aroma. Instrument parameters relating to the headspace extraction condition of volatile compounds, the GC column and the IMS unit were ne-tuned to achieve the best peak separation and signal responses. In order to develop a quantitative method, several analytical hurdles, including the usage of internal standards, and the non-linearity in signal-concentration relationship, were examined, before trialling two di erent non-linear ing methods (Boltzmann function and generalised additive models) to establish calibration curves. Performance metrics, such as goodness-of- t, precision and recovery, were thoroughly inspected. Speci c issues relating to the SHS-GC-IMS instrument, such as interfering matrix e ects caused by ethanol levels and competitive ionisation, are also discussed. e analytical capability of using SHS-GC-IMS was eventually ascertained by comparing its quantitative results with those obtained from an established headspace-solid phase microextraction-gas chromatography-mass spectrometry (HSSPME- GC-MS) method in analysing the same batch of commercial wine samples. e developed SHS-GC-IMS method was then used to construct a tripartite quality grading model of end-ferment Sauvignon Blanc wines. A total of 143 samples were submi ed to SHS-GC-IMS analysis. Using the SHS-GC-IMS data in an untargeted approach, computer modelling of large datasets was applied to link aroma chemistry via prediction models to three wine sensory quality gradings. Six machine learning models were compared, and arti cial neural network (ANN) scored a prediction accuracy of 95.4%. Despite its inherent complexity, the ANN model o ered intriguing insights on the in uential volatiles that correlated well with higher and lower sensory gradings. ese ndings could, in the future, guide winemakers in establishing wine quality, particularly during blending operations prior to bo ling. Oxidation and its impacts on wine aroma was also investigated using SHS-GC-IMS from a novel perspective: partition coe cients and partial pressure of volatile compounds in wine. Emphasis was placed on ten volatile esters and their changes during 30-day oxidation periods in both commercial wine and model wine. Physicochemical properties that can contribute to the partition behaviours of ten volatile esters, such as hydrophobicity and matrix temperature, were rstly explored. Partition coe cients were then linked to quantitative measurements to obtain partial pressures, which describes the availability of volatile compounds in the gas phase. e concept of partition coe cients and partial pressure have then been applied to a time series of aroma changes due to oxidation in commercial wines. As a follow-up study, a full factorial design was devised to inspect the impact of three common wine matrix components, namely, copper, polyphenols and ascorbic acid, on the partial pressure changes a er 30-day oxidation treatment in either full-alcohol or low-alcohol simulated wine matrices. Interesting interactive e ects between antioxidant behaviours and alcohol levels were elucidated, especially around the controversial use of ascorbic acid in winemaking. ese results can guide winemakers who wish to minimise oxidative damage to wine aroma during wine storage or bulk transport, where ullage may be present or continual oxygen ingress may be occurring. Ageing of wine can impart both positive and negative changes in aroma, depending on the duration of ageing and the storage temperature. e 24-month ageing e ects on four New Zealand wine varieties (Sauvignon Blanc, Pinot Noir, Pinot Gris, Chardonnay) were inspected using both HS-SPME-GC-MS and SHS-GC-IMS. HSSPME- GC-MS was focused on the quantitative analysis of 40 volatile compounds, which provided detailed insights into the changes in volatile concentrations during ageing. Di erent compounds respond di erently towards ageing with upward, downward, or stable trends. E ects of higher storage temperature to accelerate the loss of fruitiness of freshness was also observed. OAV values additionally revealed that not all volatiles that evolve during ageing had direct impacts on the overall aroma. e ability of semi-quantitative SHS-GC-IMS in showing the overall aroma changes during ageing was similar to that of HS-SPME-GC-MS using partial least squares discriminant analysis (PLS-DA). is highlights the potential application of SHS-GCIMS with an untargeted method to help winemakers di erentiate wines that show advanced ageing due to inappropriate storage conditions.